A mixture-of-experts transformer foundation model pretrained on diverse SEM images enables generalization across materials and outperforms SOTA on unsupervised defocus-to-focus restoration.
Materials Advances5(14), 5698–5708 (2024) 34
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Mixture of Experts Foundation Model for Scanning Electron Microscopy Image Analysis
A mixture-of-experts transformer foundation model pretrained on diverse SEM images enables generalization across materials and outperforms SOTA on unsupervised defocus-to-focus restoration.